Metadata-Version: 2.4
Name: chatcortex
Version: 0.3.0
Summary: Research framework for multi-objective synthesis and evaluation of AI agent architectures under bounded budgets
Author: Siddharth Saraswat
License: MIT
Project-URL: Homepage, https://github.com/siddharth1012/chatCortex
Project-URL: Repository, https://github.com/siddharth1012/chatCortex
Project-URL: Issues, https://github.com/siddharth1012/chatCortex/issues
Keywords: multi-objective-optimization,agent-architecture,llm-agents,pareto-optimization,architecture-search,beam-search,ai-systems-research
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: networkx>=2.8
Requires-Dist: numpy>=1.23
Provides-Extra: dev
Requires-Dist: pytest>=7.0; extra == "dev"
Requires-Dist: black>=23.0; extra == "dev"
Requires-Dist: ruff>=0.1.0; extra == "dev"
Provides-Extra: viz
Requires-Dist: matplotlib>=3.7; extra == "viz"
Dynamic: license-file

# ChatCortex

**ChatCortex** is a research-oriented framework for automated synthesis
and multi-objective optimization of LLM-based agent architectures.

It is designed as a controlled experimental platform for modeling,
generating, evaluating, and optimizing tool-augmented AI agent systems.

------------------------------------------------------------------------

## Vision

ChatCortex aims to eliminate manual agent wiring and ad-hoc prompt
engineering by introducing:

-   Formal task modeling
-   Constraint-aware synthesis
-   Multi-objective optimization
-   Exact Pareto frontier computation
-   Deterministic and stochastic execution simulation

Long-term goal:

Automated architecture synthesis of reliable AI agents from high-level
intent.

------------------------------------------------------------------------

## Architecture Overview

ChatCortex is organized into layered research components:

1. **User / Intent Layer** (future)
2. **TaskSpecification** (formal model)
3. **Synthesis Engine** (greedy / exhaustive)
4. **AgentGraph** (DAG representation)
5. **Execution Engine** (deterministic / probabilistic)
6. **Telemetry**
7. **Evaluation Harness**
8. **Pareto Optimization**

------------------------------------------------------------------------

# Core Components

## 1️. Component Metadata

Formal, immutable representation of:

-   Models
-   Tools
-   Memory modules
-   Verification modules

Each component defines:

-   Capabilities
-   Cost
-   Latency
-   Reliability
-   Privacy level

------------------------------------------------------------------------

## 2️. TaskSpecification

Defines a task formally:

-   Ordered required capabilities
-   Hard constraints (max cost, latency, privacy)
-   Multi-objective weights

Separates feasibility from optimization preference.

------------------------------------------------------------------------

## 3️. Synthesizers

### HeuristicSynthesizer

Greedy deterministic builder selecting best component per stage.

### ExhaustiveSynthesizer

Explores full Cartesian architecture space and enables exact Pareto
frontier computation.

------------------------------------------------------------------------

## 4️. AgentGraph

Directed Acyclic Graph (DAG) representation of agent architecture.

Aggregates:

-   Total cost (additive)
-   Total latency (sequential assumption)
-   Aggregate reliability (multiplicative)

------------------------------------------------------------------------

## 5️. Execution Engine

Supports:

-   Deterministic mode (structural validation)
-   Probabilistic mode (reliability simulation)
-   Fixed random seed for reproducibility

------------------------------------------------------------------------

## 6️. Evaluation Harness

Runs experiments across:

-   Multiple tasks
-   Multiple synthesizers
-   Multiple stochastic trials

Produces:

-   Average cost
-   Average latency
-   Success rate

------------------------------------------------------------------------

## 7️. Pareto Optimization

Exact multi-objective Pareto frontier computation across:

-   Cost (minimize)
-   Latency (minimize)
-   Reliability (maximize)

Provides ground-truth optimal architecture trade-offs.

------------------------------------------------------------------------

# Research Positioning

ChatCortex is intended as:

-   A systems-AI research framework
-   A controlled environment for architecture optimization experiments

It emphasizes:

-   Reproducibility
-   Formal modeling
-   Separation of concerns
-   Experimental rigor

------------------------------------------------------------------------

# Roadmap

### Phase 1 (Complete)

-   Formal modeling
-   Heuristic synthesis
-   Execution simulation
-   Evaluation harness

### Phase 2 (Complete)

-   Exhaustive architecture search
-   Exact 3-objective Pareto optimization

### Phase 3 (Planned)

-   Heuristic search (beam search, evolutionary refinement)
-   Statistical robustness analysis
-   Intent-to-task automation layer
-   Real model/tool integration

------------------------------------------------------------------------

# Installation

pip install chatcortex

------------------------------------------------------------------------

# Status

ChatCortex is currently a research framework under active development.

It is not yet a production agent orchestration library.

------------------------------------------------------------------------

# License

MIT License

------------------------------------------------------------------------

Developed by Siddharth Saraswat
